search this blog

Thursday, June 17, 2021

Balto-Slavic drift


A few years ago I began using the term "Balto-Slavic genetic drift" to describe the fine-scale genetic signal that is shared by the speakers of Baltic and Slavic languages to the exclusion of Europeans without significant Balto-Slavic ancestry.

As a result, nowadays, many people online use the term "Balto-Slavic drift" when referring to this phenomenon.

The easiest way to prove that Balto-Slavic drift exists is to run a fine-scale Principal Component Analysis (PCA) of European genetic variation with a lot of Balto-Slavic samples in the mix. Indeed, my Global25 PCA analysis does a great job of illustrating the impact of Balto-Slavic drift on the population structure of Europe both in PCA plots and mixture models (for instance, see here).

It's also possible to tease out Balto-Slavic drift with formal statistics. I showed this indirectly in a recent blog post about Greek population structure (see here). In this post I'm going to demonstrate how to explicitly and formally test for Balto-Slavic drift both in ancient and present-day samples.

To do this we need to find stats that basically split Baltic and Slavic speakers from other Europeans, such as f4(Outgroup,Test;Bell_Beaker_NDL,Baltic_LVA_BA). In this f4-stat, Baltic_LVA_BA is the ancient reference population with an unusually high level of Balto-Slavic drift, while Bell_Beaker_NDL is a fairly similar population overall in terms of ancient ancestry components, but with practically zero Balto-Slavic drift.

Note that the statistics with the most significant Z scores (>3) involve populations that speak Baltic or Slavic languages, or their neighbors who plausibly harbor significant Baltic and/or Slavic ancestry. Among the ancient, mostly Scandinavian, populations (from Margaryan et al. 2020 and marked with the VK2020 prefix), significant Balto-Slavic drift only appears in the more easterly and/or later groups from the Viking Age (VA).


Unfortunately, one of the problems with this analysis is that Baltic_LVA_BA and Bell_Beaker_NDL aren't identical in terms of their ancient ancestry proportions. For one, the latter has significantly more Neolithic farmer ancestry. No wonder then, that Greeks, who are mostly of early farmer stock, don't show a significant Z score, despite probably packing a significant amount of Balto-Slavic ancestry dating to the Middle Ages.

In the near future, as more ancient samples become available, it might be possible to find better reference populations for the job and create more accurate, finer-scaled tests.

See also...

Uralian genes

That old chestnut: Northeast vs Northwest Euros

233 comments:

1 – 200 of 233   Newer›   Newest»
Davidski said...

And if any of you geniuses out there want to point out that Uralic speakers in this analysis show a lot of Balto-Slavic drift, yeah, they do, but it's also easy to split Balto-Slavs from Uralics, unless the Balto-Slavs have a lot of Uralic ancestry. See here...

https://eurogenes.blogspot.com/2018/05/global25-workshop-2-intra-european.html

https://eurogenes.blogspot.com/2015/09/uralic-genes.html

Andrzejewski said...

I thought the Bell Beakers had more Neolithic than Balto-Slavic drifted pops.

In any case, West Euros’ Neolithic ancestry stems from GAC while BA stock is more Sourhern. Right?

Genos Historia said...

So I guess this confirms N1c came to Finland with people who had some Baltic Bronze age ancestry. Possibly people similar to Estonia Iron age.

Finland has mixed 'Estonian' & Scandinavian & Saami ancestry?

Andrzejewski said...

@Genos Historia “ So I guess this confirms N1c came to Finland with people who had some Baltic Bronze age ancestry. Possibly people similar to Estonia Iron age.

Finland has mixed 'Estonian' & Scandinavian & Saami ancestry?”

Uralic languages came from Siberia with N1c.

However, all Uralic speakers in Europe (outside of Russia) have 50% -60% Indo-European Steppe aDNA.

That includes Finns, Estonians, Saami, and even Hungarians!

Genos Historia said...

What explanation do you have for why Balto-Slavs have signature drift, but in other regions of Europe do not?

For example, why isn't there British Beaker drift in the UK? This would make sense.

Genos Historia said...

@Andrze,

Yes, Finns are Indo Europeans in disguise. It will be hard for some of them to accept this.

Davidski said...

@Andrzejewski

The farmer ancestry in Balto-Slavs is likely to be mostly from Corded Ware in that border zone between Poland, Belarus and Ukraine, but we need more ancient samples from this area to work this out, and indeed to work out the question of Balto-Slavic origins.

So far, very few samples have been published from this region from any period.

Davidski said...

@Genos

Genetic drift doesn't happen at a fixed rate. Sometimes it's much faster and more powerful, usually in populations with low effective population sizes.

https://en.wikipedia.org/wiki/Genetic_drift#Time_to_fixation_or_loss

As for the origin of Finns, it's way more complicated than them just being a mix of Estonians, Scandinavians and Saami.

Davidski said...

@Andrzejewski

I thought the Bell Beakers had more Neolithic than Balto-Slavic drifted pops.

Yep, fixed.

Anonymous said...

Some points/questions:
1) maybe instead of LVA_BA you could use LTU_BA or similar? It definately has the drift needed but it also has higher Neolithic.
2) if we go back to the Proto-Balto-Slavic population. Would you think it was a) 100% LVA_BA like and it mixed with other populations to get into modern Balts and Slavs proportions or b) LVA_BA was PBS + something else? Is there any way to test that somehow?

Davidski said...

@Parastais

I ran some stats with LTU_BA, and clearly LVA_BA and EST_BA work better.

Maybe that's because they're higher quality sample sets, but also they're definitely more extreme examples of this type of population, and you can see that when you plot them in this PCA.

https://vahaduo.github.io/g25views/#NorthEurope

I think it'd be better to use a West Euro BA group with a really high level of HG ancestry instead of Beaker_NDL, but I can't think of one. Any ideas?

In regards to your second question, I don't think LVA_BA is actually a good proxy for Proto-Balto-Slavs.

For some reason, it just has a very high level of the so called Balto-Slavic drift, probably higher than the Proto-Balto-Slavs or even Proto-Balts.

But that's just my general impression based on all of the data I've seen and logic, and I don't know how I would demonstrate it with some stats or whatever.

Matt said...

If we wanted to get around the problem of different ancestry proportions, one idea might be to "make" a population composed of individuals from different populations but who together reach the same ancestry proportion. Like say Baltic_BA is 30% Anatolian, 30% HG, 40% steppe (for instance) then take those proportions of Barcin, WHG and Yamnaya samples and place under a shared label then run stats. Would be better to do using later populations that share more drift with later people due to more recency, but you get the idea.

Davidski said...

That could be the way to go. But I'd prefer to use a real population.

In any case, we need a pop that forms a cluster with LVA_BA in terms of deep ancestry proportions, so it has to be very symmetrically related to all the main players like Barcin_N etc.

ambron said...

David, thanks for this excellent post! I suggest reading it in conjunction with this Arza's entry:

https://slavicorigins.blogspot.com/2021/05/surplus-eef-ancestry-in-modern-day-slavs.html

In this way, we see two components that characterize the Balto-Slavs - the Balto-Slavic drift and what we can call the "Slavic EEF".

Copper Axe said...

Do CWC N47/N49 have "Balto-Slavic" drift? They have a similar profile to Spiginas2 which apparently does have this drift.

Matt said...

Yeah, I can see that, particularly since if some samples have more coverage it could produce results unexpected from the ancestry proportions of the "population". So would need some testing. Though don't know if we'll ever get that ideal test population - just don't seem any pockets of HG people where such samples may have come from in West Europe. Synthetic "population" where we're going "9 Sintashta plus 1 HG" or some other synthetic setup may end up being the best we can do...

@Sam, take with a grain of salt but I think in PCA some of the drift common to West Europe tends to get rolled up into the common West Eurasia PCA, possibly because it shows some present day correlation with the underlying deeper proportions in the modern populations on which we're projecting. Certainly it looks to me on some of these PCA like MN European farmers from West Europe are not exactly placed intermediate HG and EN for example.

(How much genetic drift has proceeded since late Copper Age / late Neolithic in Europe is interesting though. One way I could imagine testing it is to compare heterozygosity (ideally at ancestral polymorphic SNPs ascertained in an outgroup) for different European regions to the expected heterozygosity on their combination of ANF:WHG:STP (including Nganassan etc for relevant populations). If a population is less heterozygous than the expected sum of its parts, that's probably due to some unmeasured drift somewhere in the phylogeny and pops with more of it would show more, while if more heterozygous, vice versa. That would help distinguish for'ex, declining South-North European heterozygosity between ancestral structure and population size... All beyond my biostatistics skills though!)

Genos Historia said...

@Matt,

I think you ran a test like that before. You tested to see how much drift and how much ancient components can explain variation in Europe in G25 PCA.

There is some drift in Northwest Europe. But not a lot.

I'm thinking balto Slavic drift comes from a small Corded Ware group who expanded.

Davidski said...

@Copper Axe

I guess that based on this they might be somehow closely related to the population in which this drift developed.

CMR_Shum_Laka_8000BP CWC_POL47-49 Bell_Beaker_NLD Baltic_LVA_BA 0.000971 3.940

CMR_Shum_Laka_8000BP Spiginas2 Bell_Beaker_NLD Baltic_LVA_BA 0.001805 6.242

But based on this (using all of the CWC_POL samples from that site), I reckon that it's their unusually high level of forager ancestry that is pushing them towards LVA_BA.

CMR_Shum_Laka_8000BP CWC_POL Bell_Beaker_NLD Baltic_LVA_BA 0.000549 2.526

Lukaszer said...

What about Av1 and Av2 samples in such test?

Davidski said...

They're not in my current dataset, but someone else here might be able to run them.

Tigran said...

When and where did I2a-DIN get incorporated into the Slavic gene pool?

Rob said...

Chimp.REF AV2 Netherlands_BellBeaker Latvia_BA 0.001498 5.287 39025 37966 707008
Chimp.REF AV1 Netherlands_BellBeaker Latvia_BA 0.000925 3.315 40086 39410 730667
Chimp.REF SZ1 Netherlands_BellBeaker Latvia_BA 0.000227 0.847 50996 50783 936164


@ Tigran

''When and where did I2a-DIN get incorporated into the Slavic gene pool?''

Probably from the outset. I.e. There were no 'Slavs' before I2a-Din

Arza said...

@ Tigran

Already in the Bronze Age, somewhere between SE Poland-W Ukraine and Hungary-Romania. Generally this haplogroup is linked to Chalcolithic and Bronze Age groups from Hungary and Romania.

Matt said...

@Sam, that was one attempt, but the better way probably to test for drift level is with a comparison to an outgroup and with direct sequence data.

Matt said...

On the subject of detecting drift and fine structure, this might interest you lot:

Part 1 of 2:

Using the new admixtools, we can generate tables of the raw f2 statistics between populations.

Here's an example for Europe and some selected ancients: Graphic: https://imgur.com/a/3OMWSow / Pastebin: https://pastebin.com/yiJQXbWw

Now using this for Neighbour Joining Trees: https://imgur.com/a/ILGpdfI

As you can see, these are really good for NJ trees, and abundantly recreate the modern fine structure of Europe, at least as well as Fst and perhaps even better. It also places the ancients I've included in quite logical places along the tree.

The disadvantage seems to be though that you do occasionally get some quite long branches. You can see in the attached that Scottish has quite a long branch from English, which reflects generally elevated values, and this might be due to real drift, but it's probably not could just be due to something in the data (which is more likely since no one has ever said before that there was a huge drift effect in Scots, and neither is this too plausible!).

This is even more of a disadvantage when it comes to ancients - I've included 12 ancient Eurasian populations in the attached table, however I did run it with a larger number of populations originally, including 8 more, but these generated very long branches that didn't match with how much the populations should be differentiated (e.g. Poland_Globular_Amphora gave a branch that was about 10x as long as Spain_MLN, for no particular reason, Germany_Mesolithic and Karelia_HG about 10x as long as IronGates for no reason, etc).

So using raw f2 statistics seems to be even more vulnerable to long branches due to artefacts etc than even fst.

However, you can get some interesting stuff out of it:

A) Plots of English vs Czech f2 distances: https://imgur.com/a/b554FqZ . Because these populations have pretty similar deep proportions, you can really see an offset in the distances between them where Czech is systematically closer to Latvia_BA (and East European neighbours) and England closer to Spain_MLN and West European neighbours.

B) Even more apparent when we make PCA based on these distances: https://imgur.com/a/0IPFUIW . Strong offset of Latvia_BA towards higher distance to English, strong offset of Spain_MLN to distance from Czech while other populations who are more evenly related to both are in between. (E.g. we can see the "Balto-Slavic vs Atlantic farmer" distinction?)

Matt said...

Part 2 of 2:

C) Now running parts of the table through a Principle Coordinates Analysis (PCoA): https://imgur.com/a/Tum8au4 .

First two images include Nganasan, Mongolia_N, China_Coastal_Neolithic_En (Boshan and Bianbian) and Cameroon_SMA, then the next one is just running PCoA on West Eurasia. It really recreates the basic dimensionality, and it surprised me that the first dimensions, with all populations, included both the West Eurasian shape and the world variation, which is unusual. Higher dimensions on this also do we'd expect them to, although there is some of what worryingly looks like it may be pseudo-dimensionality reflecting differences between processing methods/damage for ancients and moderns (maybe using transversions would limit this or something).

Running these PCoA results through Vahaduo gives some interesting and expected results as well, , though 1) it's not as "clean" PCA projection in terms of avoiding some low level components that probably aren't real and reflect quirks in the data and 2) runs with Steppe_En give slightly compressed ancestry which again probably reflects quirks in data.

Example 1: https://imgur.com/a/642eeNF
Example 2: https://imgur.com/a/oZOFYzV

(Although there are some possibly low level not real components in these fits, e.g. slightly inflated East Asian / Iran_N / Levant_C ancestry across Europe, etc., it's interesting that these are systematically areal across Europe today, and not random at all! So reflect some geographic pattern even if spurious in the literal sense.)

Hopefully interesting to some; I remember someone asking about the possibility of using f2 stats directly for analyses (without intermediation by qpAdm or qpGraph). The answer to me seems to be that you can do it and it might give higher resolution than other methods even, but it's super-sensitive to data quality, maybe much more so than many other analyses.

Rob said...

@ Arza

''Already in the Bronze Age, somewhere between SE Poland-W Ukraine and Hungary-Romania. Generally this haplogroup is linked to Chalcolithic and Bronze Age groups from Hungary and Romania.''

Which samples ? I haven't seen it come up at all. They're all other varieties of I2a,,
The closest I've seen is in a Motala -SHG. Obviously too distant to be directly relevant

Genos Historia said...

The Tollense warriors in Germany have plenty of I2a. I2a rich pops extended north of Danube in the Bronze age.

I can bet deep in Poland, there was also plenty of I2a in the Bronze age. Cousins of the Tollense warriors.

Andrzejewski said...

@Genos Historia “I can bet deep in Poland, there was also plenty of I2a in the Bronze age. Cousins of the Tollense warriors.”

Sure. But how much of this I2a stems from GAC?


Rob said...

Well, we dont know because we lack genomes. In post-CWC terms, we have an R1b-P312 and R1a-? Z280 in Choplice-Vesile culture from Poland. So far, none of the Unetice-related I2as are of GAC variety

One of the Wielbark genomes is reported to be i2a2 (GAC) lineage in the abstract (but not sure of its exact C14 date). We see a few GAC-derived I2a2 in the Vikings. Tt is present in ~ 15% of modern day north Germans.

So it lingered on, to be sure, however, the pre-medieval migrations certainly erased much I2a diversity due to the expansion of R1a-rich Slavs and R1b-rich Rhenish Germans in the Ottonian era, etc

DragonHermit said...

I'm surprised Ashkenazi Jews are that low. You'd figure they'd be higher, but I guess their mix is mostly Middle Eastern + Southern European even though most of them lived in Eastern Europe.

If this is accurate, it also seems like ancient Greeks were more southern shifted/Mediterranean, and the Slavic invasions pushed them closer to the central Balkans to populations like Albanians. They seem higher than a lot of non-Slavic places.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Check this out: https://i.imgur.com/JYivp8g.png

From Krause's new book that just came out.

PIE in Iran.

Davidski said...

Is that the same Krause the clown who claimed in a clip at Youtube that Yamnaya was from Central Asia and Germans only have 10% Yamnaya ancestry?

He seems like an awesome authority on the subject.

Kitty said...

Hi Davidski! Are you planning on bringing back the G25 store soon? I had a friend who wanted to buy coordinates but you closed the store for the time being. Also, are your modern samples still up to date or do they still need vetting?

Carlos Aramayo said...

Johannes Krause, Director of the Max Planck Institute for Evolutionary Anthropology, and journalist Thomas Trappe have published this book last 13 April 2021, untitled _ A Short History of Humanity: A New History of Old Europe_. The map is from Chapter 6: "Europeans find a language", pp. 119-136.

https://tinyurl.com/36su7hzj

Davidski said...

@Kitty

I don't know yet. I'm kind of bored with the G25 and looking at other things and how to offer them to people.

So stay tuned.

EastPole said...

So called Balto-Slavic drift has nothing to do with the origin of Slavic languages. The source of it is not-IE, EEF/HG.

You cannot date Slavic languages. Oleg Nikolayevich Trubachyov wrote about it:

“However, the question now is not that the history of Slavonic may be measured by the scale of the II to III millenniums B.C. but that we can hardly date the ‘emergence’ or ‘separation’ of proto-Slavonic or proto-Slavonic dialects from Indo-European dialects because of the proper uninterrupted Indo-European origin of Slavonic.
The latter belief is in line with the Meillet’s indication that Slavic is an Indo-European language of archaic type, vocabulary and grammar of which has not experienced shocks in contrast to, for example, the Greek (vocabulary)”

This is what probably happened in reality:

https://postimg.cc/s1gPDT9c

We know that some post Corded Ware Slavic tribes migrated south and formed Nitra culture. I am using a “willow-leaf” metal complex map because Nitra is a continuation of it, i.e. Nitra has “willow leafs” and in addition there is plenty of Slavic R1a there.

https://postimg.cc/yW6K6BxG

Notice that on PCA Fuzesabony_MBA can be a mix of CWC and Mako_EBA

https://postimg.cc/LnhhdVHT

Slavs were probably the Hyperboreans who influenced Greek religion and introduced some elements similar to Vedic.
Balto-Slavic genetic cline is the result of a long history of mixing and migrating and one cannot find Slavic homeland from this. There is no need for absurd theories about a small group of drifted Balto-Slavs suddenly, miraculously, expanding in the Middle Ages.

Davidski said...

@EastPole

Two questions:

- Why do Balts and West and East Slavs carry the highest levels of Balto-Slavic drift?

- Why does Balto-Slavic drift suddenly show up in ancient DNA wherever/whenever Slavic languages are also first attested?

Rob said...

@ East Pole
Theyre ancient Balto-Slavs and such para-groups
but modern Slavic derives from a very recent expansion. We dont even need to pretend to date languages, the near mutual intelligibility of such languages to this day speaks volumes, not to mention the very clear case of demographic changes between 400 and 700 AD; now backed by directly by aDNA

ambron said...

David, Balanowski answers your questions well:

"The article carefully assumed that the assimilated substrate could be represented mainly by the Baltic-speaking populations. Indeed, archaeological data show a very wide distribution of the Baltic groups before the start of the resettlement of the Slavs. The Baltic substrate in the Slavs (though along with Finno-Ugric) was also revealed by anthropologists. The genetic data obtained in this work - both in the genetic dependence graphs and in the proportion of total genome fragments - indicate that the modern Baltic peoples are the closest genetic neighbors of the East Slavs. Moreover, the Balts are linguistically the closest relatives of the Slavs. And we can assume that by the time of assimilation, their gene pool was not so different from the gene pool of the Slavs who began their extensive resettlement. Therefore, if we assume that the Slavs settling in the east assimilated mainly the Balts, this may explain the similarity of modern Slavic and Baltic peoples to each other, as well as their differences in relation to the surrounding non-Slavic groups of Europe."

Davidski said...

@ambron

So Slavs first assimilated Balts pretty much everwhere north of the Balkans, and then they took their Baltic admixture with them into the Balkans?

VJ said...

Could the reason for the drift be that the balto-slavs descent from a population that went through a long period isolation in which they experienced a genetic bottle-neck, that might have been very severe (or may even be several such population reduction events), after which there was a sudden and rapid expansion (around 400AD)?

This is my first comment here, I'm really an amateur, so I might be shooting in the dark here.

ambron said...

David, the Slavs already had such an admixture before migration, as shown by the Central European genomes from the Iron Age.

"Moreover, the Balts are linguistically the closest relatives of the Slavs. And we can assume that by the time of assimilation, their gene pool was not so different from the gene pool of the Slavs who began their extensive resettlement".

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

I'm agnostic towards Krause's book and theories. Not saying I agree with it just thought it was interesting. He thinks Beakers spoke Celtic. One thing I thought was interesting was he talked about how the oldest evidence for use of the wheel came from FBC, and that FBC had a lot of technical innovation. Fair to say the consensus is now that the wheel did originate in the European Neolithic.

Synome said...

@Davidski

There are theories that the Thracian and Dacian peoples are partially descended from Baltic or Baltic like migrants based on linguistic links.

Is it possible to test this hypothesis yet with Balto Slavic drift
measures, or do we need more relevant samples?

Arza said...

@ VJ
Could the reason for the drift be that the balto-slavs descent from a population that went through a long period isolation in which they experienced a genetic bottle-neck, that might have been very severe (or may even be several such population reduction events)

That's exactly what has happened. That population was a WHG-rich group that survived in the vicinity of Hungary practically until the Bronze Age.

Davidski said...

@VJ

Long isolation isn't necessary for this sort of genetic drift to happen.

All we need is a population derived from a few founders expanding rapidly over an area of low population density.

So considering that LVA_BA has so much of this signal, then this has to be a signal which developed in a Corded Ware group expanding somewhere south of Latvia.

Tigran said...

I2a-Din is L621 I believe. How is that related to the Yamnaya clade of I2a? Has that survived among Slavs or anybody els?

Davidski said...

@Synome

Are there any Thracian or Dacian samples available?

Curtis said...

@Davidski I've been wondering something concerning haplogroups and the tracing of their histories.Supposing you have the proper tools and a good DNA database and you run a comparison of the populations which have the haplogroup of your interest( from those with higher to low frequency and from different geographies) with those that don't but are "relative" populations ,is it somehow possible to find other unique genes(?), mutations(?) that you could somehow relate to the expansion of the specific haplogroup's carrier population?
If something of that sort is possible , it could maybe help "extract" more haplogroups indirectly from ancient samples or find out about complete replacements.There's probably stuff I'm missing since I'm no geneticist but it's intriguing to think.

Synome said...

@Davidski

I was hoping you'd know that answer to that question, hah. I'll look around and see. What do we have on the iron age Balkans?

Rob said...

“ Are there any Thracian or Dacian samples available?”

There’s BGR_IA

Andrzejewski said...

@Arza “
That's exactly what has happened. That population was a WHG-rich group that survived in the vicinity of Hungary practically until the Bronze Age.”

That Hungary Hg thing lol…it’s contagious: first Samuel Andrews brought up it, later Genos Historia featured Hungary, Serbia and Poland HG in his Tollensee Valley video, and now you.

Next Lazaridis is going to write up an essay about “Hungary BA HG”.

Matt said...

@Sam/Genos, comment below which follows on from our discussion upthread as to whether drift has been a stronger force in Eastern or Western Europe overall, and also linking to my above exercise with the f2 statistics.

Set of plots comparing f2 distance to euclidean distance under the G25, for a selected populations, first distance from Yoruba, then distance from Belarusian, then distance from Jordanians: https://imgur.com/a/pYbl204

There is a very strong match between a power relationship where the f2 distance predicts the G25 distance. So the G25 is very strongly relating to and capturing the f2 distance. The overall tree structures on both have very similar local and global structure (though different magnitudes to each).
(This is interesting if we've ever wondered how distance in these PCA relate to the formal f2 distance that forms, through various additive combinations, the backbone of all the formal f-statistics analyses done in adna. As a general aside, it might be useful to have a look and cross check how well distances match between PCA and f2 stats, as a check to whether a PCA is representing a faithful model of the real f2 stats!).
However, when we look at the fine-scale there are a few distances, where anything above the line of best fit has a higher G25 distance than would be expected for f2 distance, and beneath the line has a lower distance:
- In the Yoruba comparison, the G25 distances (y-axis) tend to be slightly higher in East Europe from a given latitude/geographic isolation, but in the direct f2 comparison (x-axis), they're often as high comparing Western European populations.
- Likewise looking at distance from Belarusian and distance from Jordanian, again the Global 25 is massively capturing everything at the important, high scale level, however at a fine scale we see that G25 (y-axis) estimates a higher distance between Belarusians and Southern European groups than is the case in f2 distance, while predicting lower distance between Belarusians and NW European groups (and Finns and Basques!)
(Put another way in a fine scale context, if we went by the G25 distance, we'd expect Belarusians to be slightly closer to NW Europeans than to Bulgarians and Romanians, but the f2 statistic is about the same.
Obviously a prediction that distance from Yoruba is similar in NE and NW European populations would then suggest that the cumulative drift away from the African origin has been similar in both sets of populations.

So basically I'm not saying the f2 distance here is necessarily exactly right (although it *is* calculated directly on the sequence data), but just trying to demonstrate there can be some slight difference between PCA like G25 and other measures, so it is difficult to tell. It's a very fine scale question, so getting a definitive answer on it would be difficult. (Not a slam against the G25, the G25's great and absolutely I think covers most all the Global scale differences that won't vary much no matter what samples are included and even covers most relatively local scale differences. Just that we getting into very fine scale differences in this topic, and so these are more difficult for a PCA to absolutely cover within absolute tolerances).

Rob said...

Thraco-Dacian and late Balto-Slavic probably had no special relationship, if anything contact with indo-Iranian is more likely
The distinctive feature of BGR_IA is recent Anatolian admixture.

Arza said...

@ Andrzejewski

It's my years-old hypothesis and Sam has nothing to do with it.
https://twitter.com/GerberDniel2/status/1400010765445353475

Davidski said...

So Balto-Slavic drift correlates closely with a major branch in the Indo-European family tree.

And it comes not from a subset of the Corded Ware population that expanded rapidly (you know, like the Corded Ware people did) but an extremely isolated and highly unusual group from the Carpathian Basin that spoke...what???

Yeah, OK.

Leper said...

What do you think about Turkish_Northwest? For a long time it's been speculated that Turks from around northwest of Turkey have some Slavic ancestry mixed into their "native" admix.

bce said...

I-L621 isn't connected to GAC or Yamnaya, it has the highest diversity in Western Germany and probably originated there in the neolithic:
https://www.yfull.com/tree/I-L621/

Rob said...

“ The distinctive feature of BGR_IA is recent Anatolian admixture.”

Although might be mediated via Greece

Rob said...

@ Romulus

“ He thinks Beakers spoke Celtic”

Statements like these are pretty useless. Academics should at least try to understand some nuance & dynamism, or otherwise not really bother

Matt said...

On the (slightly off topic) f2 statistics plots upthread, made some more examples of the kind of graphics and analyses that can be created with them in PAST, and saved them as PDFs in order to be easily zoomed into (so the labels don't just become a blur).

Links here in a Pastebin: https://pastebin.com/cz2A4Rvs

The f2 statistics on a set of modern populations look to capture essentially all the structure that we could expect to see.

Matt said...

@arza: may be of interest in respect of Daniel Gerber's research -

https://agi.abtk.hu/en/news/news - "Prehistoric mass grave from the coast of Lake Balaton "

"Preliminary results indicate an extremely exciting genetic makeup for the individuals of Kisapostag culture as it signalises recent immigration into the region. Their connection to the BBC can be demonstrated [Fig.4.], although it seems to be minor, and they rather represent a population previously unseen in Prehistoric Europe."

"The research continues, final results will be presented as the backbone for the PhD dissertation of Dániel Gerber, publication and detailed description are coming up this year."

"Dániel Gerber – Anna Szécsényi-Nagy – Viktória Kiss"

Seems lots of I2a. I'm unsure if this is a new article or not as seems undated on the page.

Matt said...

Re my above comment, here are some more results that may relate to the site of Balatonkeresztur and potentially to Gerber's adna research there:

Poster from SAA 2019 - http://bakota.net/wp-content/uploads/2019/SAA%20Posters/Giblin%20et%20al_SAA%20Poster_2019_FINAL.pdf

Ongoing project page on ResearchGate - https://www.researchgate.net/project/Balatonkeresztur-Bronzkor-Kisapostag-Meszbetetes-Balatonlelle-Baden

Samples are from the so called "Encrusted Pottery Culture" - https://www.sciencedirect.com/science/article/abs/pii/S0305440307000775

Matt said...

Daniel Gerber's research into the apparently I2a rich population of Balatonkeresztur does seem to be current according to what I can tell from here - https://agi.abtk.hu/en/research/current-research-projects and the link to Viktória Kiss' Moment Project - http://mobilitas.ri.btk.mta.hu/?lang=en

Genos Historia said...

@Andrze,

My view, is the rich WHG ancestry in the Bronze age comes from Lithuania not Hungary.

Lithuania hunter gatherers are a better reference than Southeastern hunter gatherers. We don't have DNA from Poland hunter gatherers, except one sample, they might be the actual source.

The Tollense's warrior's fit as 30% derived from Narva Lithuania hunter gatherers. Some of them 40%. Some Hungary BA guys also fit as 40% derived from Narva.

ambron said...

This WHG and Z280+ rich Nitra population is reportedly originally comes from southern Poland. Such words were reportedly spoken at the conference where these samples were presented for the first time.

Andrzejewski said...

@Genos Historia “ My view, is the rich WHG ancestry in the Bronze age comes from Lithuania not Hungary.

Lithuania hunter gatherers are a better reference than Southeastern hunter gatherers. We don't have DNA from Poland hunter gatherers, except one sample, they might be the actual source.

The Tollense's warrior's fit as 30% derived from Narva Lithuania hunter gatherers. Some of them 40%. Some Hungary BA guys also fit as 40% derived from Narva.”

Davidski on a previous thread shot down the misleading notion that one of the differences between Eastern and Western Euros is due to Baltic HG.

So I guess it didn’t stick around to contribute to Balto-Slavs.

What happened to this HG rich pop?

Arza said...

@ Matt
Thanks for the links! Exciting times ahead!

Next two sentences are even more important:

Our task turns to hard mode from this point as __recovering unknown genetic sources__ is a tough nut to crack. Fortunately, increasing number of archaic genomes eases the placement of certain groups, we “only” need the correct __mathematical models to make assessments__ on origin for example.

Unknown hunter-gatherers confirmed. As predicted years ago.

Arza said...

@ Davidski

Would it be possible to add Vatya to the G25? Or at least the HUN_MBA_Vatya_o2:RISE479 (861756 SNPs in Reich's dataset). They're from the site (Erd) mentioned in the poster linked by Matt.

PS
Chimp Estonia_BA.SG Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00204 0.000634 3.22 0.00127 211740
Chimp Lithuanian Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00163 0.000535 3.05 0.00227 229905
Chimp Latvia_BA Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00157 0.000577 2.72 0.00648 228279
Chimp Belarusian Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00129 0.000532 2.43 0.0152 229905
Chimp Ukrainian_North Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00130 0.000534 2.43 0.0151 227581
Chimp Ukrainian Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.00120 0.000531 2.25 0.0245 229905
Chimp Norwegian Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.000935 0.000525 1.78 0.0747 229905
Chimp English Hun_MBA_Vatya.SG Hun_MBA_Vatya_o2.SG 0.000728 0.000527 1.38 0.167 229905

Arza said...

HO:
Cameroon_SMA_pu Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00101 0.000314 3.21 0.00135 415931
Cameroon_SMA.SG Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00114 0.000325 3.51 0.000447 418633
Cameroon_SMA.DG Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00102 0.000327 3.13 0.00173 417196

1240k:
Cameroon_SMA_pu Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00137 0.000291 4.71 0.00000243 744462
Cameroon_SMA.SG Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00148 0.000294 5.02 0.000000512 751011
Cameroon_SMA.DG Hun_MBA_Vatya_o2.SG Netherlands_BellBeaker Latvia_BA 0.00134 0.000300 4.46 0.00000808 747957

Arza said...

https://www.ebi.ac.uk/ena/browser/view/PRJEB43715

Genome of a Middle Holocene hunter-gatherer from Wallacea

Much remains unknown about the population history of early modern humans in Southeast Asia, where the archaeological record is sparse, and the tropical climate is inimical to the preservation of ancient human DNA. Here we report the first ancient human genome from Wallacea, the oceanic island zone between the Sunda Shelf (mainland Southeast Asia and the continental islands of western Indonesia) and Pleistocene Australia-New Guinea (Sahul). We extracted genetic material from the petrous bone of a young hunter-gatherer woman who was buried 7274–7078 calibrated years before present at Leang Panninge, a limestone cave in South Sulawesi, Indonesia. The individual shares most genetic drift and morphological similarities with present-day Papuan and Australian groups, yet represents a previously unknown divergent human lineage that branched off around the time of the split between Papuan and Australian populations ~37 thousand years ago5. We also describe Denisovan- and deep East Asian-related ancestries in the Leang Panninge genome and infer their large-scale displacement from the region today.

Genos Historia said...

Thanks for showing this article on Bronze age Hungary DNA, Matt.

Prehistoric mass grave from the coast of Lake Balaton
https://agi.abtk.hu/en/news/news

80% Y DNA I2a. The rest R1b, R1a.

Genos Historia said...

@All,

Life like reconstructions of Yamnaya and other prehistoric people. There's four Yamnaya ones.

https://www.deviantart.com/philipedwin/art/Yamnaya-Boldyrevo-I-Kurgan-I-Grave-1-873552929

I don't know who is making these reconstructions. But it'll be cool to see faces given to ancient people who have been mysterious for a long time.

Maybe by 2035, they'll finally make a documentary on Indo Europeans.

Genos Historia said...

The reconstructions come from this guy. I will link him in my youtube videos.

https://www.deviantart.com/philipedwin

Copper Axe said...

Those reconstructions are terrible.

First of all some of those are Srubnaya rather than Yamnaya, and the pigmentation is way off.

South Asians with rs16891982 C/C and rs1426654 A/A are lighter than these "Yamnayans", where based on current samplrd the vast majority had at least one derived copy and plenty had two. From what I've seen with european people with one derived copy of rs16891982, such as Russians, is that their skin tone is still fairly "white" so to say.

So I would not include these reconstructions in your video Genos.

Matt said...

@arza / Sam, thanks, yeah looks interesting esp. the mentions of "exciting genetic makeup", given Gerber's tweet of his "sidequest" to look into HG rich outliers and the picture we've slowly built up (ever since BR2 was released in the dawn of adna era way back in 2014 and we talked about unusual pattern of the connections and other findings emerged).

Also yeah, the y-dna mention of "Paternal lineages of the Kisapostag culture seem to be novel in the region" and the I2a frequency looks high - it looks like they must've got at least 15 male samples (as that's the minimum number that could allow them to state R1b was present in 6.66% samples), though some of these look to be relatives (which I guess makes y-dna frequency estimates less "independent"). (Likewise it seems like male+female samples must be at least 19 to get the mtdna proportions).

Davidski said...

@Arza

Scaled

HUN_MBA_Vatya_o:RISE479,0.132035,0.140143,0.087869,0.066215,0.069859,0.018128,0.011751,0.01223,0.036201,0.013668,0.005521,-0.014087,0.01888,0.038534,-0.029044,0.000265,-0.003651,0.012162,0.006662,0.00075,0.011979,0.000124,-0.021692,-0.099773,0.007305
HUN_MBA_Vatya:RISE247,0.133173,0.153345,0.034318,-0.013243,0.054472,-0.007809,0.012926,0.001385,0.032315,0.031891,-0.00065,0.011839,-0.018285,-0.006331,-0.018594,-0.005038,0.011474,-0.000633,-0.000628,0.012131,0.008235,0.012118,-0.017131,-0.002169,-0.018322
HUN_MBA_Vatya:RISE480,0.122929,0.144205,0.037335,0.008075,0.040007,-0.01506,-0.00705,0.005307,0.015748,0.035172,0.000487,-0.001199,-0.002676,-0.005367,-0.014658,0.008221,0.005867,-0.004307,-0.002137,0.002501,-0.010107,0.003586,-0.003451,-0.016026,-0.00946
HUN_MBA_Vatya:RISE483,0.132035,0.15436,0.041106,-0.002261,0.051086,-0.001673,-0.002585,-0.003231,0.016362,0.015672,0.017538,0.005095,0.004162,-0.004817,-0.021308,-0.013126,-0.005215,0.006334,0.00817,-0.005628,0.001248,0.005935,-0.003821,-0.015544,-0.004431
HUN_MBA_Vatya:RISE484,0.126344,0.139128,0.048271,0.027132,0.045239,0.004741,0.012926,-0.000462,0.016362,0.02934,0,0.009292,-0.006095,-0.004954,0.008279,-0.003845,-0.010561,-0.001014,0.008296,0.005878,0.000624,0.005564,-0.004314,-0.017954,-0.00012

Raw

HUN_MBA_Vatya_o:RISE479,0.0116,0.0138,0.0233,0.0205,0.0227,0.0065,0.005,0.0053,0.0177,0.0075,0.0034,-0.0094,0.0127,0.028,-0.0214,0.0002,-0.0028,0.0096,0.0053,0.0006,0.0096,0.0001,-0.0176,-0.0828,0.0061
HUN_MBA_Vatya:RISE247,0.0117,0.0151,0.0091,-0.0041,0.0177,-0.0028,0.0055,0.0006,0.0158,0.0175,-0.0004,0.0079,-0.0123,-0.0046,-0.0137,-0.0038,0.0088,-0.0005,-0.0005,0.0097,0.0066,0.0098,-0.0139,-0.0018,-0.0153
HUN_MBA_Vatya:RISE480,0.0108,0.0142,0.0099,0.0025,0.013,-0.0054,-0.003,0.0023,0.0077,0.0193,0.0003,-0.0008,-0.0018,-0.0039,-0.0108,0.0062,0.0045,-0.0034,-0.0017,0.002,-0.0081,0.0029,-0.0028,-0.0133,-0.0079
HUN_MBA_Vatya:RISE483,0.0116,0.0152,0.0109,-0.0007,0.0166,-0.0006,-0.0011,-0.0014,0.008,0.0086,0.0108,0.0034,0.0028,-0.0035,-0.0157,-0.0099,-0.004,0.005,0.0065,-0.0045,0.001,0.0048,-0.0031,-0.0129,-0.0037
HUN_MBA_Vatya:RISE484,0.0111,0.0137,0.0128,0.0084,0.0147,0.0017,0.0055,-0.0002,0.008,0.0161,0,0.0062,-0.0041,-0.0036,0.0061,-0.0029,-0.0081,-0.0008,0.0066,0.0047,0.0005,0.0045,-0.0035,-0.0149,-0.0001

CrM said...

Philip Edwin's reconstructions are pretty good imo.
I sometimes like to give a shot at facial reconstruction myself. Here's Barma Grande 2 for example.
https://i.imgur.com/lKgu4V2.png

Matt said...

To get a really good pipeline for reconstructions in future, with the advances in GAN generation of faces, I reckon they could:

1) train one of these Stylegan models on faces of present-day people, but incorporate genetic PCA into the underlying model, to allow the model to project faces beyond present day genetic ancestry variation.
2) ... train on hirisplex scores too.
3) ... finally train on scans of present day x-rays of facial skeleton and real faces as an additional constraint.

Combining these, which is feasible with present day tech, should allow to get pretty accurate best guesses of faces. But it's a question of whether anyone with the money and technical skills wants to invest in doing that.

Eventually will probably be able to do that in motion.

I don't think those reconstructions are too bad but there are probably things wrong with them from taking present day references and warping to fit.

Arza said...

@ Davidski

Thanks!

Distance to moderns is expectedly gigantic, but quite telling at the same time.

Distance to: HUN_MBA_Vatya_o:RISE479
0.11415488 Polish:Polish7
0.11506040 Polish:Polish37
0.11507614 Polish:Polish27
0.11624926 Russian_Smolensk:RUS_Smol345
0.11682559 Lithuanian_VZ:LTG-664
0.11689071 Belarusian:Belarusian11
0.11698179 Ukrainian:Ukrainian16
0.11723398 Polish:Polish6
0.11737741 German:German75
0.11820753 Russian_Voronez:RussianVoron103
0.11831259 Belarusian:Belarusian9
0.11837618 Russian_Smolensk:RUS_Smol312
0.11851525 Lithuanian_VA:LTG-603
0.11859593 Lithuanian_PA:LTG-173
0.11865544 Belarusian:Belarusian7
0.11878994 Ukrainian:UKR-1992
0.11881221 Lithuanian_SZ:LTG-1112
0.11888778 Polish:Polish17
0.11889835 Belarusian:Belarusian6
0.11924798 Belarusian:Belarusian15
0.11957666 Belarusian:Belarusian1
0.11977137 Lithuanian_VA:LTG-372
0.11981350 German:German13
0.11994202 Russian_Smolensk:RUS_Smol304
0.12013562 Lithuanian_VA:LTG-598

Arza said...

Target: HUN_MBA_Vatya_o:RISE479
Distance: 4.4852% / 0.04485218
53.6 Baltic_LVA_BA
20.4 Iberia_Southwest_CA
11.4 HUN_ALPc_Tiszadob_MN
7.2 ITA_Grotta_Continenza_Meso
6.8 FRA_Nouvelle_Aquitaine_Meso
0.6 IRN_Tepe_Abdul_Hosein_N

BTW That "normal" Vatya is very close to the non-Baltic_BA part of the Slavs:

Target: Ukrainian
Distance: 2.2914% / 0.02291367
55.4 Baltic_EST_BA
44.6 HUN_MBA_Vatya

Target: Belarusian
Distance: 1.7422% / 0.01742152
62.6 Baltic_EST_BA
37.4 HUN_MBA_Vatya

Target: Polish
Distance: 2.0176% / 0.02017581
52.0 Baltic_EST_BA
48.0 HUN_MBA_Vatya

Davidski said...

@Arza

This individual is an outlier for a reason. In other words, this person was not native to the Carpathian Basin, not even in an outlier kind of way.

Probably a descendant of recent migrants from somewhere in the east/northeast.

EastPole said...

Target: HUN_Fuzesabony_MBA:I20772
Distance: 2.3933% / 0.02393284 | R4P
25.8 POL_Chlopice_Vesele_Culture:I6537
20.2 HUN_MBA_Vatya_o:RISE479
17.0 Corded_Ware_POL:N47
16.8 Corded_Ware_POL:pcw361
13.2 Corded_Ware_POL:pcw350
7.0 Corded_Ware_Proto-Unetice_POL:RISE431

https://postimg.cc/LnhhdVHT

Rob said...

Yeah , im not sure what the fuss is about ? I called bullshit on the 'CWC 70% wipeout' theory whilst the rest of you were slobbering like schoolgirls back in 2015 :)
There's no magic refuge, these WHG, I2a rich come range from France to Ukraine, all of which could have made their way to the Carpathian basin.

Matt said...

Where this drift came from and where these outliers came from is a very uncertain thing still. I think the "isolated Carpathian/Central HG" hypothesis is definitely in the running (and strengthened I think by the academic work mentioned) but it does have to explain why the richest level identified so far survived or was found around the Baltic, exactly where late HG were also known, and how or why late HG would survive in Central, where they're not really as known ( though far from implausible - Blatterhohle, Wartburg). It does seem that the drift is most associated with the enriched HG ancestry, so I think that's a question for if we were thinking it was associated with a Corded Ware group that was not originally HG enriched - it does seem to be more complicated to say it wasn't associated with extra HG then it was than it just being associated with enriched HG to begin with.

Matt said...

Are we sure the Vatya outlier (RISE479, Erd, I2a) is projecting OK? Looks to have a *very* Balto-Slavic rich position on clines, possibly more so than LVA_BA.

E.g. graphics from Vahaduo's PCAs: https://imgur.com/a/hDcglKD

Very extreme position that doesn't seem to be explicable by enriched HG ancestry alone.

I'll do some modelling to compare later. The easier way to get a feel for it is to compare the real sample to a reference model comprised of references without Balto-Slavic drift and then see where the residual plotting differences arise.

Theplayer said...

Unrelated but do you know what happened to the Dzudzuana paper? @Davidski

Simon_W said...

Johannes Krause is barely 41 years old.

https://en.m.wikipedia.org/wiki/Johannes_Krause

The German mainstream media consider him to be the young shooting star of DNA science. :D

ambron said...

David, we already have a surprisingly large number of these migrants from the northeast.

Simon_W said...

@Carlos Aramayo

"Johannes Krause, Director of the Max Planck Institute for Evolutionary Anthropology, and journalist Thomas Trappe have published this book last 13 April 2021, untitled _ A Short History of Humanity: A New History of Old Europe_. The map is from Chapter 6: "Europeans find a language", pp. 119-136.

https://tinyurl.com/36su7hzj"

From the book description:

"Genetics has earned a reputation for smuggling racist ideologies into science"

Really? Did it?

"but cutting-edge science makes nonsense of eugenics and “pure” bloodlines. Immigration and genetic exchanges have always defined our species; who we are is a question of culture, not biological inheritance."

Has this actually been proven that biological inheritance is irrelevant? And has anyone proven that cultures are in no way influenced by the DNA of the people? (I'm asking this rhetorically, in case anybody wondered.) Also, Eugenics cannot be "disproven" by cutting-edge science. Because eugenics aims at breeding mankind, in order to improve its genepool. How can a political choice like this be "disproven"?

Simon_W said...

Oops, re: Krause as a shooting star; apparently the expression is a false friend. In German it means someone who is new and exciting, very fahionable too. But according to dict.leo.org it has a more negative meaning in English. :D

Matt said...

Quick attempt at a RISE479 model and then differences from model: https://imgur.com/a/5da40AJ

Using Vahaduo's distance difference mode, the differences between the base and the target are systematically shifted towards BS cline populations.

Here's another attempt at modelling: in this one I used *all* population averages (both those before and after it), but every time it returned a population that I knew had "Balto-Slavic" drift/cline or a modern, I removed it:

The first model I got that wasn't including any "BS Cline populations" - https://imgur.com/a/DL2xjYC

It seems weirdly expected that this model is majority HUN_CA plus LTU_Narva with some minority ROU_Meso. Makes good geographic sense. However there's also a weird preference for TKM_BA that makes less sense.

Lowest fit I could get it to take using population averages strictly from Europe and strictly before or at the same time as it: https://imgur.com/a/wgOzwym

Got some weird Yamnaya/CWC aversion that's probably a PCA artefact.

Davidski said...

@ambron

So can you give a plausible alternative explanation for this sample's genetic structure and clear outlier status?

Obviously, we both know that his ancestors weren't hiding in a cave somewhere in the Carpathian Basin.

In any case, he doesn't look very Balto-Slavic to me. More like from some unusual parallel line from the east.

Arza said...

@ Matt

Remove PC24 completely or transplant it from the nearest sample (I1502).

I11600 seems to behave the same way.

Check PC2/3/24 https://vahaduo.github.io/3d/g25/

WE PCA coords from 2017:
Vatya:RISE247,-0.0212,-0.025,-0.013,0.0075,0.0024,0.0118,0.0045,-0.0007,0.0017
Vatya:RISE479,-0.0475,-0.0098,0.0159,-0.0093,0.0041,-0.0082,-0.0039,-0.0018,-0.0066
Vatya:RISE480,-0.0245,-0.0156,0.0004,0.01,0.0001,0.0086,0.0017,0.005,0.0022
Vatya:RISE483,-0.0275,-0.0184,0.0005,0.0008,-0.003,0.0027,-0.0111,0.0027,-0.0027
Vatya:RISE484,-0.0322,-0.0124,0.0027,0.0032,0,0.01,0.0009,-0.0148,0.0011

Desdichado said...

"Has this actually been proven that biological inheritance is irrelevant? And has anyone proven that cultures are in no way influenced by the DNA of the people? (I'm asking this rhetorically, in case anybody wondered.) Also, Eugenics cannot be "disproven" by cutting-edge science. Because eugenics aims at breeding mankind, in order to improve its genepool. How can a political choice like this be "disproven"?"

No, quite the opposite. Behavior is 80% genetic. Political affiliation is now heritable than height, etc. Culture is downstream from genetics.

ambron said...

David, for my taste RISE479 has too few steppe accents for a migrant from the east. I suspect that RISE479 is one of the last Mohicans of the Carpathians that Hungarian and Slovak geneticists talk about.

Davidski said...

@ambron

You're confusing two issues.

One, there was a very WHG-rich population, or even populations, in the Carpathian Basin that survived there well into the Bronze Age.

Two, there were all sorts of outliers living in the Carpathian Basin during the Bronze Age, some of whom more or less resembled Corded Ware and Bronze Age samples from the East Baltic.

Obviously, the latter were migrants into the Carpathian Basin.

ambron said...

David, that's what I'm talking about. This individual lies on the WE PCA on the opposite side to CWC, that's why does not look like a migrant.

Matt said...

If the "Balto Slavic drift" is separable from HG ancestry, another unusual thing is that there is this sample ROU_C_o that takes it but seem to have no steppe ancestry and this sample that seems to want very little of it. That there are samples rich in HG but who seem to take no or little steppe ancestry, if another argument beyond the Baltic_BA populations that it originates from drift in a HG group.

Davidski said...

@Matt

You must be confusing excess HG ancestry with Balto-Slavic drift.

https://1.bp.blogspot.com/-oRY5HAPpAnY/YNGXc5-DPNI/AAAAAAAAKGE/qAL71JWiRukuKyFfest8dFWiwccz25TPQCLcBGAsYHQ/s1200/Vahaduo%2BGlobal%2B25%2BEurope%2B1%2BPCA.png

Davidski said...

@ambron

He doesn't but another one does.

But that's not a problem, because we can't expect all of the migrants that came to the Carpathian Basin to have been from the same area and population.

Arza said...

@ Matt

https://i.postimg.cc/z8ZJyc9K/Balto-Slavic-quintessence.png
https://i.postimg.cc/Bs1SQ0wJ/Balto-Slavic-quintessence2.png

This time it's not calculated from coordinates. It's Baltic_BA with masked SNPs that have alleles common with CWC. ~25000 SNPs.

Matt said...

@davidski, try other PCAs and also consider the distance difference between ROU_C and other samples with similar deep proportions. The sample is very shifted towards Balto-Slavic Cline *relative* to other samples with same proportions (I believe ROU_C has rich HG ancestry and EEF, no steppe, if I remember it right, it is very shifted towards "East" even in such a PCA as you posted, compared to other samples with similar proportions). You may be confused.

Davidski said...

@Matt

I'm not confused, you are.

If an ancient sample actually has Balto-Slavic drift, rather than just something related at some level, then it'll behave like a Balto-Slav in all significant dimensions of PCA that pick up this drift.

Matt said...

@arza, interesting way to replicate. West Eurasia PCA, not Global 25?

Davidski said...

@Arza

I may have mentioned this before, but I'll mention it again: that cline doesn't represent what you think it represents.

That is, it's not a cline that runs from modern Slavs to our ghost forager ancestors with the original Balto-Slavic drift.

It's actually a cline that once ran from our farmer ancestors to our WHG ancestors, but it was shifted east by steppe admixture.

In other words, the steppe admixture came so suddenly, that it shifted almost all populations to the east, largely preserving that cline, but mostly only in dimension 2.

Arza said...

@ Davidski

Re: ROU_C_o
https://i.postimg.cc/tp7M8tTm/Vahaduo-Custom-PCA-8.png

It's not Matt who is confused.

Matt said...

@Davidski, so rather than the Balto-Slavic drift happening in the Corded Ware horizon (as I think you said upthread) it is "related on some level" to drift in a HG / HG+EEF population?

Davidski said...

@Arza

You're both very confused.

But I'm hoping that more aDNA will clear up this nonsense.

Rob said...

There seem to have been populations with high HG until quite late.

But they all lie on slightly different clines, and most became extinct during Iron Age -> Medieval periods

As I previously mentioned, the I2a2a1a of Vatya guys is similar to Welzin, which in turn finds the closest link to British Neolithic. This lineage has not surfaced in Baltic-HGs, for ex.

Davidski said...

@Matt

Well, duh, late CWC groups in East Central Europe carried a lot of local ancestry, including excess forager ancestry.

Matt said...

So a drift process in a HG rich CWC which happened after ROU_C causes it to be in a different position on PCA? Also I thought you were initially saying the drift happened in a CWC population which was not particularly rich in HG, is that not the case?

Davidski said...

@Matt

You thought wrong.

I always made the point that Spiginas2 was the earliest example of a genetic Balto-Slav, and that the process that led to this sort of genotype was rapid.

Matt said...

So your view is that LVA_Baltic_BA = proto-Balto-Slavic, and has undergone sudden drift effect which, due to PCA, shows up in a phantom way in HG sample like ROU_C_o?

Matt said...

So in you paradigm, we'd never expect to see a population which shows "normal" Corded Ware proportions but an equal or more intense level of Balto-Slavic drift than LVA_BA (which I think is what "ColdMountains" believes existed?)

Davidski said...

No, I don't believe that LVA_BA = proto-Balto-Slavic.

As I once explained here, I think that LVA_BA doesn't exactly resemble Proto-Balto-Slavs or even Proto-Balts. It's way too extreme and unusual.

In fact, I believe that the Proto-Balto-Slavs carried less of the so called Balto-Slavic drift and less HG ancestry.

Matt said...

As a quick check of where we are, just trying to state the three main hypotheses (and this is not an attribution of originality, just labelling them according to who seems to be most strongly and consistently arguing for them).

Hypothesis A: the arza hypothesis: Balto-Slavic drift happened at some point in the history of a specific late surviving HG population which became incorporated into a later Bronze Age composite with steppe and EEF that forms a pole in the ancestry of people today.

Hypothesis B: the ??? hypothesis (not purely owned by anyone): Balto-Slavic drift happened in an isolated EBA composite of Corded Ware and HG, who were initially not particularly drifted. This then forms a pole in the ancestry of people today. Any presence of BS drift in earlier HG populations w/out CWC ancestry is simply an illusion of PCA.

Hypothesis C: the ColdMountains hypothesis: BS drift happened in a CWC population formed purely off the basis of GAC+steppe. This population is the true common ancestor of Balto-Slavs. Any finding in PCA that contradicts and suggests LVA_BA forms a pole in ancestry and is co-ancestral to Slavic population with Carpathian Bronze Age, or that early outliers with high HG ancestry had Balto-Slavic drift - all just inexplicable illusions of PCA.

Varying degrees of "Wholly from HG" to "Wholly from CWC".

Of these I find A + B plausible, prefer A very slightly. Find C lacking in any evidence to support it right now (other than I guess some qpAdm?).

If any of these are misstated, please restate.

I guess your hypothesis is somewhere between B and C Davidski.

Matt said...

That last bit probably a slight misstatement given what you actually posted directly above.

Coldmountains said...

@Matt

I not claimed that and agree with Davidski here. Balt_BA is both in terms of archaeology and genetics not directly ancestral to Slavs or even Balts. They represent a peripheral Para-Balto-Slavic group, which by time was replaced by more southern Balto-Slavic-related groups. Also if you would look closely at the Balt_BA samples you would see that LTU_BA is already much less HG and more EEF shifted than LVA_BA and EST_BA so there was for sure variation of EEF/HG ancestry among Balto-Slavic related groups. Groups even more in the south like in Belarus or Ukraine should using common sense be even slightly more EEF shifted since the Bronze Age (EEF groups were much more numerous here than in the Baltics which seems to be populated mainly by HGs before the arrival of CWC).

Also if Balto-Slavic drift is from the Carpathian basin then why the non-outliners lack any shift towards Balto-Slavs and cluster with Spaniards/South French and not even with South Slavs. His pseudo-Balto-Slavic shift is just the result of super high WHG ancestry of a similar type which Proto-Balto-Slavs picked in Poland, West Ukraine or Belarus. The same is true for the Tollense samples except for WEZ56 and other Hun_BA samples who lacked any Balto-Slavic Y-DNA and only get shifted towards Balto-Slavs when they are unusually rich in WHG. Samples from the same region lower in WHG always lacked this shift and had so far not in one case Y-DNA related to Balto-Slavs

Davidski said...

My hypothesis is pretty simple and logical.

Corded Ware groups expanded rapidly over low population density areas of East Central and Eastern Europe.

This resulted in admixture with the locals but also rapid drift and thus late Corded Ware populations like Spiginas2.

These late Corded Ware populations eventually formed post-Corded Ware groups like LVA_BA and the Proto-Balto-Slavs, but the story may be more complex in the case of the latter, or at least the Proto-Slavs, based on some ancient DNA that I've seen.

I'm really not sure how this idea that the Balto-Slavic drift originated in foragers from the Carpathian Basin got started, but it sounds like something that Carlos Quiles would come up with.

Matt said...

BS drift correlated with HG ancestry in all PCA, and pretty plausible that HG groups experience strong drift, since they all have the lowest population sizes. Seems plausible enough to me; I don't understand the counterarguments that it's not plausible for drift to happen in a small HG population, or that are seemingly vehemently against absolutely any contribution of Carpathian Bronze Age populations to later people of the same broad region.

EastPole said...

@Davidski

„No, I don't believe that LVA_BA = proto-Balto-Slavic.

As I once explained here, I think that LVA_BA doesn't exactly resemble Proto-Balto-Slavs or even Proto-Balts. It's way too extreme and unusual.

In fact, I believe that the Proto-Balto-Slavs carried less of the so called Balto-Slavic drift and less HG ancestry.”

Could you explain your theory of Proto-Balto-Slavs, because I am also “confused” and don’t follow you.
Notice that there are some doubts as to whether the Balto-Slavic should be understood as a common ancestor (i.e. a common period in the development of Slavic and Baltic languages). It is only agreed that of all Indo-European branches, no other branch can be considered more closely related. There are many theories.

Could you illustrate your theory on a PCA like for example this drawing from the comments on Arza’ blog:

https://i.postimg.cc/yBpxQjgb/isbspspb.jpg

This theory explains linguistic relations of Slavic with other languages and the presence of Slavic R1a in other populations.

https://postimg.cc/Cdx2fpz0


I hope your theory will explain it too.

Coldmountains said...

@Matt
Name me please a plausible archaeological culture from the Carpathian basin contributing to BA Baltics and East Europe. Or which kind of Y-DNA is shared between Hungary_BA and Balto-Slavs?

Davidski said...

We're talking here about drift that very specifically defines a branch of the Indo-European tree.

It didn't happen in any forager population. It happened during the Indo-European expansions.

You can't project such drift back in time, because it didn't exist at one point, so you'll just be pushing older samples into these sorts of spaces, and obviously they have to go somewhere.

Rob said...

@ cold Mountains

“ Balt_BA is both in terms of archaeology and genetics not directly ancestral to Slavs or even Balts. They represent a peripheral Para-Balto-Slavic group, which by time was replaced by more southern Balto-Slavic-related groups. ”

This is incorrect .
There was no population replacement in the Baltic; not since after CwC
Modern Balts are simply more southern due to returning Balts & admixture with neighbours
These LBA

Baltic LBA were northern balto-Slavs; pretty much as simple as that.

Davidski said...

@EastPole

I've got a surprise for you, or not if you've been following what I've been saying here.

Balts and Slavs are both from a population intermediate between Baltic BA and something a lot more southern from just east of the Carpathians, and you'll see this confirmed with ancient DNA, along with all of the right Y-DNA clades.

Coldmountains said...

@Rob

What do you even want to say that? It was already by archeology+linguistics proposed that Balts arrived in the Iron Age from the southwest and genetic pretty much confirm that. You had several immigration waves from the southwest and of them brought Baltic languages and replaced some related Para-Balto-Slavic languages spoken before.

Matt said...

@coldmountains, in general I think it's plausible from a broad region are ancestral to later ones, and I don't think the burden is on me to try and prove this through a painstaking archaeological continuity which you will contest. Y-dna is heavily structured and can expand or contract with subpopulation dynamics of a broad continuous genetic cline.

Matt said...

"A lot more Southern", huh? (Any comment there, ColdMountains?)

Rob said...

@ cold Mountains

“ What do you even want to say that? It was already by archeology+linguistics proposed that Balts arrived in the Iron Age from the southwest and genetic pretty much confirm that”

Nope
I repeat- The great population shift occurred after CWC . Then by ~ 1000 BCE the “fortress people”/ striated pottery arrived. Then there’s continuity until the MAs , excepting local movements
You need to read up about the “returning Balts” (mercenaries of Attila and Avars from the south)

Davidski said...

Yeah, a lot more southern, but who isn't more southern than Baltic BA?

These weren't Minoans or anything like that. Still in the North Euro range, broadly speaking.

Coldmountains said...

@Matt

Most archaeologists and experts link Balto-Slavs with the Trzinec-Komarov complex, which originated to a large extent from East Polish CWC subgroups so can you find a link between Trzinec-Komarov and the Carpathian Basin? For a theory to be logical you not just need not just fancy PCA experiments but also archaeology, uniparental markers, other genetic analyzes (qpAdm,..) and common sense to support it and so far I don't see much pointing in this direction. Using Occam's razor it is not more likely that Proto-Balto-Slavs originated near Spiginas2 somewhere between East Poland, West Ukraine and Belarus around 2000-2500 B.C than outside of CWC in population lacking any Balto-Slavic related Y-DNA and belonging to cultures with at best dubious archaeological links to the Baltics and East Europe? Most linguists also place Balto-Slavic between Germanic and Indo-Iranian and the Pre-IE substrate of Balto-Slavic is close to that of Germanics and Celts but distant to that of Albanian and Greeks ,what points to Balto-Slavs picking these non-IE terms from GAC/TRB groups in CWC.

Arza said...

@ Davidski

"Recent effective population size in Eastern European plain Russians correlates with the key historical events"
https://www.nature.com/articles/s41598-020-66734-y

Can you take a plot from this study and mark a spot when that sudden bottleneck has happened between 5 and 4 kya (166-133 generations ago)?

Matt said...

So what kind of proportions do you reckon? If 50:50 Baltic_BA like + "more southern" wouldn't be too hard to quickly estimate what we're talking about here.

Davidski said...

@Arza

Nope, not really interested.

Btw, I just sent you an email. Don't fall off your chair when you see it.

Davidski said...

@Matt

For northern Slavs it's gonna be about 30-40% Baltic BA-like if we're talking LTU_BA, and much less if LVA_BA.

And for Balts I reckon the opposite and then some, so about 70-80% Baltic BA.

Rob said...

The other perhaps more important source is ongoing admixture with a series of neighbouring groups over the epochs, esp. when there's Y-DNA and at least broad archaeological continuity (Gimbutas, Shchukin, Bliujiene). You might see it from Luzatian, Pomoranian, Wielbark, Poles, etc). Sure, there might be from as far as the Carpathian region (Otomani culture-> Chernolis-> Baltic) but it would be pretty minor.


Matt said...

Ok, so with that as a figure, if LTU_BA = 1, and Ukrainian / Belarusian = 0.35, then regression predicts that 0 is as follows:

Imgur: https://imgur.com/a/kbVnVoZ Pastebin: https://pastebin.com/Lf5pFWRd

Basically a population already like a present day Slovakian or Southern Polish person.

(If you had 50:50 LTU_BA, X into Ukrainian/Belarusian, then it would be more like a Bosnian person: https://imgur.com/a/vf3q8s6).

@Rob, I don't think everyone is necessarily suggesting that it is plausible for a direct migration from Carpathian Bronze Age, but more by isolation-by-distance geneflow across over a thousand years.

Davidski said...

Basically a population already like a present day Slovakian or Southern Polish person.

Yeah, something like that, but more eastern.

Matt said...

I guess that would imply that populations like broadly Southern Poles already existed somewhere before formation of Balto-Slavic (as a language)? Somewhere...? But not in Carpathian region. Or is this too much extrapolation.

EastPole said...

@Davidski
“I've got a surprise for you, or not if you've been following what I've been saying here.

Balts and Slavs are both from a population intermediate between Baltic BA and something a lot more southern from just east of the Carpathians, and you'll see this confirmed with ancient DNA, along with all of the right Y-DNA clades.”


When PIE split there was only Cucuteni–Trypillia culture and variants of CT east of the Carpathians. Is it what you are suggesting?
I would also like to fall off my chair, why only Arza?

Davidski said...

@EastPole

CT was long gone by that time. Take a look at some maps of Late Neolithic/Early BA east Carpathians. CWC and Yamnaya are all over this area, then Unetice, Nitra etc.

@Matt

That sounds about right, but let's wait and see.

Matt said...

OK, let's see.

One thing that this sort of two way model would have as an advantage over HUN_Vatya+LVA_BA model like Arza's upthread is that HUN_Vatya+LVA_BA can't quite cover the required amount of Steppe ancestry, under G25, so you'd need to add some complexity to cover that. And we generally think sources that could "top up" steppe ancestry tend to progressively have too much Turan + East Asian related ancestry over time to be able to do that.

On PCA: https://imgur.com/a/YBG8Qys

Although maybe this whole line of argument is not that significant. IDK.

Mauri said...

Yeah, the Finns are mixed Estonian' & Scandinavian & Saami, but the result wouldn't be the same using present-day populations. Every population today if more mixed and drifted. No, the Finnish N1c didn't came from Estonia, only to some extent, let's say maybe 30%. During the Iron Age large areas in Fennoscandia, Baltic area and Northern Russia were populated by N1c. Let's not make this as a linguistic drivel before we have ancient genomes to watch. This is quite a cimplex issue; who spoke which language 4000 years ago. Mostly we can only say that today we have simultaneity to a certain extent thanks to the historical continuity, but wedon't exactly know WHEN and WHERE this continuity started.

Carlos Aramayo said...

@Simon_W
"...Has this actually been proven that biological inheritance is irrelevant? And has anyone proven that cultures are in no way influenced by the DNA of the people? (I'm asking this rhetorically, in case anybody wondered)..."

I think in the future the linguists and philologists will think they have the last word and Genetics will be used by them only as "auxiliary means" to adjust the "timing" of the migrations or changes. But Geneticists can have the oportunity to explain their views as a separate "discipline" or as separate "streams".


Genos Historia said...

Maybe EastPole is right about CT.

Maybe Corded Ware admixed with CT which explains the much lower WHG in Slavs, Bronze age Poland compared to Baltic BA.

Genos Historia said...

So Central-East Europe in Bronze age maybe there was hodge podge of different types of Europeans.

Some with high WHG, low Steppe.
Some with low WHG, high Steppe
Some high WHG, high Steppe.
Some low WHG, low Steppe.

In Western Europe there was more of a simple two way cline, of decreasing Steppe ancestry but similar farmer ancestry.

Andrzejewski said...

@Genos Historia “ In Western Europe there was more of a simple two way cline, of decreasing Steppe ancestry but similar farmer ancestry.”

Nope. In WE the WHG went down a while EEF went up following Bell Beaker.

“ Maybe Corded Ware admixed with CT which explains the much lower WHG in Slavs, Bronze age Poland compared to Baltic BA.”

Neither. CT were Balkan Farmers + some Iron Gate HG remnants. They language must’ve been different than GAC while was either Cardial Pottery language or WHG or had different linguistic communities living side by side. The non-IE substrate would’ve been different than in Germanic had it been CT related.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Seems like Baltic CWC was ancestral to multiple other western CWC groups.

Davidski said...

Nothing supports that statement. Especially not the paternal markers.

Baltic CWC was basically all R1a. Western CWC had R1b, R1a and a lot of I2.

In fact, Baltic CWC received a lot of admixture from more southerly CWC groups, and this is how it got more farmer ancestry.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Yeah you're right, that was a poor conclusion.

What do you make of the ancient samples on this branch?
https://www.yfull.com/tree/I-Y28222/

It has samples in Mesolithic Serbia, Baltic HG, Hungary Neolithic, 1 from Czech CWC, 1 from German CWC, and 1 from Morkin. I thought the CWC samples came from Baltic HG but they are closer to the Hungarian Neolithic samples.

Rob said...

@ Rom


The Czech CWC is different branch- I2a2a2
"CHE"" refers to Switzerland - i.e. Bichon

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

@Rob

I was talking about id: I6696 CZE 4300-4850 YBP

ambron said...

So it looks like my intuition hasn't failed me. I have said many times that the best candidate for a proto-Slavic genome is a genome similar to the Slovak one, because such a genome lies at the very center of Slavic genetic variation.

Rob said...


Looks like quite a bit of structure in Baltic CWC & BA groups

On PCA some look like Yamnaya ('early baltic") , some have Narva/WHG admixture, others more Farmer

For further characterisation, can see F4 correlations which teases out steppe vs WHG vs Farmer proportions

Can try differentiate HG admixture through a series of Stats too (e.g. Iron Gates vs Narva, with usual limitations in mind

Ric Hern said...

I'm a bit confused about the possible migration route of R1b L51 descendants towards the West. +-Southern Poland towards the West along the Alpine foothills and then down the Rhine to the Low Countries and then backtrack as Bell Beakers or Southern Poland down the Elbe to the Low Countries and from there up the Rhine as Bell Beakers ? Or up the Vistula to Northeastern Poland and from there either via the Notec to the Elbe or a Maritime route from Northeast Poland to Denmark ?

Rob said...

@ Rom
nice, snuck in the 'South Asia' paper. An I2a1 individual, unlike the other Czech_CWC_o, this dude acquired steppe admixture.

EastPole said...

@Rob
“Looks like quite a bit of structure in Baltic CWC & BA groups”

Yes, a lot of structure, huge area. You cannot find a dot from which Balts and Slavs expanded. We have to go by Y-DNA and culture in addition to autosomal.

https://postimg.cc/47yWHLV3

Simon_W said...

@Genos Historia

Continued from the previous thread,

"Well, it seems ancient DNA is disproving the idea protoVillonova in Italy is from Urnfield."

So far we've only got one Protovillanovan genome, a female from northern Abruzzo. If she should be taken as typical, then the Protovillanovans indeed are not from Urnfield, because she shows close similarity to the Bronze Age and Iron Age of the Western Balkans. But maybe she was not typical; being inhumated she deviates from the cremation rite so typical of the Protovillanovan culture. And her location, close to the Adriatic sea, is predisposed to show close ties to the Dalmatian coast on the other side of the Adriatic. Apart from that, we've got one Villanovan individual - who's similar to the Latins. And as a likely representative of the Sabellic branch of Italic, we've got the individual from Boville Ernica, who is rather a member of the Sabellic Hernici than a Latin. But, genetically he's similar to the Latins, too. Overall, though, there's still little genetic evidence referring both to the Protovillanovans and to the Sabellic Italics.

"Maybe their similarities to Urnfield, are due to them following the same cultural trends as their cousins in Central Europe. This would make sense considering they had just arrived in Italy 500-600 years before."

That's impossible, there were definitely fresh cultural impulses from outside of Italy responsible for the development of the Protovillanovan culture. There are technological and ornamental similarities that cannot arise due to similar internal evolutionary trends alone, e.g. the shape of helmets or the aquatic bird motif.

"Did this study get DNA from Apennine culture? Because I thought all Bronze age Italy samples had Indo European ancestry?"

The samples from Regina Margherita BA (located in Latium) must be from the Apennine culture (Proto-Apennine to Apennine proper), if they date to 1600-1400 BC, as the wikipedia article on the Grotte di Collepardo suggests. According to the paper, their most recent samples date to 1600 BC, which would be at the border between the Proto-Apennine and the preceding Grotta Nuova facies. I guess you're right that the Bronze Age samples of this new paper all show steppe ancestry. But there are some from previous papers, which have little to no steppe:
1. ITA_Remedello_BA:RISE486; going by the date and location, must be a representative of the Polada culture, an EBA culture of northern Italy, in particular north of the Po.
2. ITA_Olmo_di_Nogara_MBA:S18130 and ITA_Olmo_di_Nogara_MBA:S18132; again going by date and site, these must belong to the Terramare culture, an MBA to LBA culture of the Po Valley, predominantly south of the Po.

So, judging from this limited available data, neither the Polada culture, nor the Terramare culture were IE.

However, ITA_Broion_EBA and ITA_Broion_BA do show steppe ancestry. They are from the province of Vicenza, in the Veneto. The main sites of the Polada culture are to the west and southwest of this province, so as far as I can tell, it's not really part of the Polada culture. But I don't know what the local culture is called.

Simon_W said...

At Broion we can see that most of the steppe ancestry was already there during the EBA; it increased only a little up to the MBA:

Target: ITA_Broion_EBA:BRC010_018
Distance: 5.1245% / 0.05124518
70.4 TUR_Barcin_N
18.4 Yamnaya_RUS_Samara
11.2 WHG

Target: ITA_Broion_BA:BRC007_019
Distance: 3.9387% / 0.03938676
67.0 TUR_Barcin_N
22.0 Yamnaya_RUS_Samara
11.0 WHG

Target: ITA_Broion_BA:BRC003
Distance: 3.4840% / 0.03484025
63.8 TUR_Barcin_N
23.4 Yamnaya_RUS_Samara
12.8 WHG

Target: ITA_Broion_BA:BRC002
Distance: 3.0185% / 0.03018485
62.8 TUR_Barcin_N
24.8 Yamnaya_RUS_Samara
12.4 WHG

On the other hand, ITA_Remedello_BA:RISE486, from the EBA Polada culture to the west of Broion:

Target: ITA_Remedello_BA:RISE486
Distance: 4.6741% / 0.04674113
82.2 TUR_Barcin_N
17.8 WHG

And, Olmo di Nogara MBA, from the MBA Terramare culture right beside the Po:

Target: ITA_Olmo_di_Nogara_MBA:S18132
Distance: 3.4077% / 0.03407695
80.8 TUR_Barcin_N
10.6 WHG
8.6 Yamnaya_RUS_Samara

Target: ITA_Olmo_di_Nogara_MBA:S18130
Distance: 5.2298% / 0.05229752
82.4 TUR_Barcin_N
14.2 WHG
3.4 Yamnaya_RUS_Samara

Only marginal steppe ancestry there. There has been argument among archaeologists as to whether the Terramare culture is descended from the Polada culture or represents a new wave of immigrants from around Hungary. The genomic data suggests continuity with the Polada culture.

Simon_W said...

There's some Terramare influence in the pottery of the LBA Sub-Apennine culture of Tuscany, Umbria and the Marche. Since the Terramare culture looks genetically non-IE, some might think that this must be the origin of the Etruscans. The problem with this view is just that the two pure Etruscans that we've got, don't show a pull into the direction of Olmo di Nogara MBA. Instead, compared to the Italics, they have reduced WHG - but not reduced Steppe. This points to admixture from the western Balkans, from Dalmatia to, maybe, Northwestern Greece. In other words, from a place closer to Lemnos than to Etruria. It will be interesting to see more Villanovan samples, and how they compare to the Etruscans. Also more Etruscan samples, to be sure.

The MBA Apennine culture seems to be Italic, judging from the samples at Regina Margherita BA. More or less contemporary with the Apennine culture was the Tumulus culture of central Europe. However, the Apennine culture can in no way be derived from the Tumulus culture. They didn't bury their dead below tumuli in the first place. I'd say they were a mix of Italian Bell Beakers with additional central European ancestry that trickled down the peninsula with bronze technology, coming from northeastern Italy. In the literature you read that Polada influence can be seen in EBA central Italy. Now, apparently this is unlikely to have been non-IE Polada proper, but I'd guess more a little from to the east of Polada, where Italic-like IE people like the ones from Broion resided. In any case, some influence from EBA groups of Southern Germany and Austria, like Singen and Straubing, did reach EBA Northern Italy.

Guy said...

Hi Folks, No aDNA but cool article. https://www.nature.com/articles/s41586-021-03675-0#Fig2 "Pleistocene sediment DNA reveals hominin and faunal turnovers at Denisova Cave" with Janet Kelso, Svante Paabo and Matthias Meyer.

Rob said...

“ Bronze age Italy samples had Indo European ancestry?"”

Technically a false statement , because IE is a language group not a type of ancestry

Genes don’t speak ! (Well actually they do modulate speech :))

Rob said...

@ Simon W


How is RISE-486 is from the Polada culture ?
Aside from the fact that it bears no cultural affinity with it, its doubtful Polada culture had even began by 2200 BC

What appears to have occurred in Northern Italy is;
- arrival of some Beaker males from Central Europe ~ 25/2400 BC
- gradual dissolution of local warrior cultures (e.g. Remedello, Spilamberto)
- further migration from Bavarian-Swiss post_BB groups forming enclaves
- continuity of multiple cave burials in other regions
- genesis of Polada culture (as a new symbiosis around Grada), with scant burial rite (odd skull deposited here & there within settlements)

Unfortunately, the data from Broion don't really tell us how the transformation happened, because the cave was re-used for millenia, and the handful of dated samples don't really give us a diachronic picture.


It might be that western Indo-European languages link back to Urnfield, or had remained in contact until that period.
- Italics -> italic peninsula
- Lusitanian: linking into the Atlantic LBA
- various para Celtic languages in mainland central-western Europe
- final formation of proto-Germanic in the urn-cremating Jastorf culture (although it obviously had its own line of development prior to that)

Genos Historia said...

@Rob,

Those Bronze age Italians look like ancestors of Latins, so they probably spoke Indo European.

I suppose they do have less Steppe ancestry than do Latins. Maybe Italic languages came from a second wave into Italy from Urnfield. But.....Steppe ancestry this far south in Italy brings into question the idea Urnfield is source of Italic.

Carlos Aramayo said...

Regarding genes and speach, did you see this?

"Being Anglo-Saxon a matter of language and culture, not genetics"
June 23, 2021

https://tinyurl.com/xafapawh

"‘Who were the Anglo-Saxons?’ is a question that has been debated for many years—and one that researchers from Simon Fraser University and the University of Sydney may have helped resolve. By exploring a new line of evidence—the three-dimensional analysis of skeletal remains—researchers have found that Anglo-Saxon identity had more to do with shared language and culture than shared ancestry.Although their origins can clearly be traced to a migration of Germanic-speaking people from mainland Europe between the 5th-7th centuries AD, the number of individuals who settled in Britain is still contested, as is the nature of their relationship with the pre-existing inhabitants of the British Isles, most of whom were Romano-British..."

"Plomp and her colleagues found that between two-thirds and three-quarters of early Anglo-Saxon individuals were of continental European ancestry, while between a quarter and one-third were of local ancestry. When they looked at skeletons dated to the Middle Anglo-Saxon period (several hundred years after the original migrants arrived), they found that 50-70 per cent of the individuals were of local ancestry, while 30-50 per cent were of continental European ancestry."

Rob said...

@ Genos

I guess MLBA time period could fit for Italic, with its links to Alpine post-Beaker groups

Ric Hern said...

p-Celtic, p-Italic and p-Greek vs. q-Celtic, q-Italic, q-Greek...The q's seems to be at least linked to Middle Bronze Age expansions. Proto-Gaelic, Proto-Latin, Proto-Mycenaean. p's were more Urnfield related.

Ric Hern said...

Some Middle Bronze Age Swords can be linked from Britain to Greece. Mycenaean, Wietenburg, Helversum, (Wessex) Cultures even Ireland...

Genos Historia said...

@Carlos,

Thanks for sharing news on upcoming Anglo Saxon ancient DNA. I read about it earlier on Survive the Jive's channel.

It is really funny.

They will come to the conclusion genes=/=ethnicity no matter what the DNA data shows. I will have fun with this paper on my youtube channel when it is published.

Rob said...

well it does say 66-75% of early Anglo-Saxons were migrants. But we have to agree that its not all about genes to show we're woke and acknowledge that the overall affair is indeed a cultural process

Copper Axe said...

"Being Anglo-Saxon a matter of language and culture, not genetics"

Anglos are too predictable lmfao.

Also, is this an ancient DNA or a craniometric study?

Genos Historia said...

Oh, yeah, it is ancient DNA!

Here is a reddit link to their abstract. It will be a big study

https://www.reddit.com/r/anglosaxon/comments/jrgngt/upcoming_archaeogenetics_study_the_anglosaxon/

Tom said...

"Anglos are too predictable lmfao."

What is that supposed to mean?

Andrzejewski said...

@Genos Historia “ Oh, yeah, it is ancient DNA!

Here is a reddit link to their abstract. It will be a big study

https://www.reddit.com/r/anglosaxon/comments/jrgngt/upcoming_archaeogenetics_study_the_anglosaxon/”

From the article there was an 80% replacement of native Celtic gene pool.

It’s not that different from Dutch Beaker replacement of the Stonehenge farmers 4,000 years ago.

However, the article prefers to highlight the assimilation of Brythonic/Cymru gene pool into the English.

But an 80% replacement rate vindicates how come the Celtic substrate in English is virtually non-existent.

EastPole said...

@Davidski

‘CT was long gone by that time. Take a look at some maps of Late Neolithic/Early BA east Carpathians. CWC and Yamnaya are all over this area, then Unetice, Nitra etc.”

Was Nitra like HUN_Gava_BA:I20771?

Target: HUN_LBA:I20771
Distance: 2.2276% / 0.02227606
34.6 POL_TRB:N19
28.0 Corded_Ware_Baltic:Spiginas2
15.6 Corded_Ware_Baltic:Plinkaigalis241
8.8 Corded_Ware_POL:N47
7.2 Corded_Ware_Baltic_early:Gyvakarai1_10bp
3.2 Corded_Ware_Baltic_early:Plinkaigalis242
2.6 POL_BKG_N_o1:N22

Carlos Aramayo said...

@Genos

Yes, it sounds a little strange to hear of "craniometric" studies nowadays. And anyway, it seems the other aDNA study you mention is not carried out by the same people from Sydney University, Australia, but from Max Planck Institute.

Rob said...

Copper axe - btw Dutch are Anglo-Saxons

Davidski said...

@EastPole

Nitra is quite different, with more steppe ancestry.

Copper Axe said...

@ Tom

"What is that supposed to mean?"

You know for decades certain academic circles, especially in the anglosphere have been fighting tooth and nail to "debunk" the Anglo-Saxon invasion/migration narrative for decades.

To no surprise then that their reaction to evidence which shows that the writings of Bede were not hogwash is this:

"Being Anglo-Saxon a matter of language and culture, not genetics"

Rather than "Oh Anglo-Saxons did migrate from the continent, and it was a very significant migration that changed everything on the island - from language, to culture, to genetic makeup."

That statement is not something you can actually derive from the skulls of people either in my opinion.

It is also kinda silly to say because from what we can tell Germanic lawcodes often had pretty strong preferential treatment for their Germanic people.

Under early Frankish law, Romans were legally akin to second class citizens for example. Another example would be that when the Anglo-Saxons migrated away, the continental Saxons began what was essentially a caste system as they gained hegemony over other Germanic peoples.

Bloodlines were for sure an important thing to these people, we can deduce that from history pretty easily. To make the statement that being AS was all about culture rather than ethnic origin - on the basis of a quarter of skulls being of local stock - is funny, and really falls in line with the denialist "consensus" within Academia regarding the Anglo-Saxons in the Anglosphere (UK, US/CAN and apparently AUS/NZ).

They can talk about settlement and brooches continuities for as long as they want, and add woke phrenology to their arguments, but genetics is making it clear they were absolutely wrong lol.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

@Tom

That was a great quote by Copper Axe

LOL

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Anglo Saxons get too much credit for a nation the Normans built

Genos Historia said...

@Carlos,

So I guess there are two Anglo Saxon studies in the works.

@Andrze,
"But an 80% replacement rate vindicates how come the Celtic substrate in English is virtually non-existent."

Good point. Hopefully the real researchers will explore stuff like this in the study.

Genos Historia said...

@Romulus,

Normans are given enough credit.

Normans created England but the English ethnicity was created by Anglo Saxons.

I do wonder if there is Norman French geneflow in England which can explain the slight 'southern' pull there today compared to Irish.

Genos Historia said...

By 'real researchers', I mean ones who don't have a political agenda. I hope they seriously look at Anglo Saxon DNA and connect it to language, archaeology, history, etc.

Andrzejewski said...

@Genos Historia “ Normans created England but the English ethnicity was created by Anglo Saxons.

I do wonder if there is Norman French geneflow in England which can explain the slight 'southern' pull there today compared to Irish.”

Anglo-Saxons, Danes and Normans - all products of the Nordic Bronze Age.

And that possible Sourhern pull you mentioned might be responsible for the English being a tad “swarthy” or Mediterranean compared to the Irish, who have slightly higher WSH element.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

@Genos

Why do you think there would not be significant Norman French geneflow? When they conquered the Isles they purged the nobility and replaced it with their own. Norman surnames are still disporportionately found in the British upper classes:

https://qz.com/301150/this-is-the-proof-that-the-1-have-been-running-the-show-for-800-years/

ambron said...

If I understand David correctly, the Northern Slavs inherit most of their genome from a population similar to the Slovaks, which however was formed somewhere further in the east through the absorption by the early CWC of the local EEF population with a large share of the WHG.

However, we need to see this remoteness to the east realistically. This could not take place further than on the eastern slopes of the Carpathians, because this is where the WHG range ended. Further there was only the hybrid HG, similar to the SHG.

http://darkheritage.blogspot.com/2018/09/indo-european-genetics.html

Davidski said...

@ambron

The link you posted looks like garbage.

Matt said...

Above mentioned study is based on basicranial shape. I don't think there's anything controversial in it - multiple distinct Germanic groups, migrated to Britain as Angles, Saxons, Jutes and formed Anglo-Saxons initially settled as a fairly dense (but already porous) core, then other ppl married into the group as the Roman trade network ceased to function, and probably by 200-300 years (Middle Saxon) or about 10-15 generations after initial migration the majority of ancestry in Anglo-Saxon settlements may not be from migrating people. All plausible. See if more direct genetics back it up. If I remember rightly it, this is what Nick Patterson's upcoming genetic work roughly also finds though.

ambron said...

David, do you have any better interpolation map to show the eastern extent of the WHG?

Davidski said...

@ambron

Well, I could make a better one, but why?

There's no evidence of any SHG-like population contributing to the Balto-Slavic genotype.

Obviously, the populations that were involved were early Corded Ware, which was very similar to Yamnaya, and farmer groups like GAC, usually rich in WHG.

Heyerdahl said...

@All

Keep in mind that the AS study is not out yet. The final product may very well be different from the abstract seen 7 months ago. 80% genetic replacement? I have my doubts. Max Planck loves ostentatious statements.

EastPole said...

@Davidski

“Nitra is quite different, with more steppe ancestry.”

Like HUN_Fuzesabony_MBA? Or more like CWC? Fuzesabony was R1a-Z280. Nitra is also R1a, but which clades? Do you know?

Davidski said...

@EastPole

Nitra generally has more steppe. It's very similar to Srubnaya, late CWC, and also to early Bell Beakers.

There's Z280 in Nitra. Don't know beyond that.

Coldmountains said...

@Coldmountains
sorry for asking (just very curious) are these "East Carpathian" samples older than Spiginas2 or contemporary with Balt_BA? Is Spiginas2 the oldest sample with BS drift you are aware of?

Matt said...

@Heyerdahl, maybe, but I would say that, reading the whole abstract, they talk about 80% replacement in *early* Anglo-Saxon graves then more incorporation over time *within* the South and East of England by the Middle Saxon period (200-300 years later). They amount of this they leave shadowy (but from what I can recall from twitter "previews" of either this or a paper by Reich lab, takes the proportions in the Middle Saxon period or later down to lower "Germanic" proportions).

I don't think this is inconsistent with the evidence people see from names and suchlike, where there is a prepondence of more British names among the dynastic elites of the more dominant kingdoms in the middle to later period, which tended to be more in the west of England later in the Middle AS period (Wessex, Mercia), like Cerdic and (probably) Penda. Probably what I guess would happen is that Angles and Saxons and Jutes could settle in a dense fashion on "Saxon Shore", but incorporated more people as the societies spread west. They probably had to, in order to thrive and succeed (and if they'd refused to do so they probably would've been far less successful as a culture than they were). Then there's more genetic influx, but the influx is not associated with Celtic content words in the language, just more like incorporating Celtic names as a father-to-son connection linking sons to established British patrilocal elites and place names, which again kind of makes sense if that's linked to legitimating prior geographic claims.

Davidski said...

@Coldmountains

Younger than Spiginas2.

Matt said...

@arza, as another alternative to identify the "Balto-Slavic" drift using formal stats, what about the following strategy:
Using Vahaduo, it seems that ROU_C_o:GB has about the same proportions of HG:ANF ancestry as DEU_Wartberg_MN:KH150422.

2 way Model: https://imgur.com/a/di4ODc6

I've included a big batch of EN+MN+LN populations that I could remember having possibly high HG samples among them; there are some distance outliers as I've not sifted out some steppe ancestry LN samples. ROU_C_o is not one of them I think, although it may have some small Steppe_Eneo related ancestry in a 3 way model, if this doesn't simply reflect a more eastern HG origin: https://imgur.com/a/mzgSpvC

This shows up quite noicely in the distance differences between the two: https://imgur.com/a/O5xULLw

So potentially that being the case, it may be a simpler way to detect Balto-Slavic drift (or at least a part of it that dates to pre-LNBA) rather than trying to match LVA_BA, by using the following stat: f4(ROU_C_o:GB, DEU_Wartberg_MN:KH150422; X, Y), where X and Y are either two test populations or an outgroup and a test population.

The problem with this is that of course you may end up with very few SNPs, so unfortunately it may not be practical, but I thought it might be worth putting it out there as a way to avoid trying to do a more complex three-way match between LVA_BA and some other populations that may not even exist.

Another strategy would be taking a broader set of MN and HG samples and making a "pseudo ROU_C_o" composite population from them, then running the f4 stat. Though I think that would be less reliable.

(Longer lists of distance difference - https://imgur.com/a/UHeNDr1 . This distance difference mode is a fairly efficient "Balto-Slavic cline finder", using only real samples. Although the Vatya outlier actually doesn't have a very strong showing in this, which is unexpected to me - https://imgur.com/a/ru0Ex7w)

«Oldest ‹Older   1 – 200 of 233   Newer› Newest»